Neural Network Trading & Investing
Technical Analysis -
There must be over 1000 different technical indicators created, some dating back over many decades. What is common between almost all of them is that they interpret differently the same stock price information that is available to everyone. That is the open, high, low, close and volume for whatever period and stock you analyze. Yes, some filter, some smooth, some repaint and some show candlestick patterns. But in the end, it’s just an interpretation of the same data. So how does a trader or investor gain an edge using technical analysis? Every stock trades differently and every stock also deviates from its norm in changing market conditions. You must have a strategy that takes into account the six different kinds of markets. This is a trade plan requirement for determining trade entries and more importantly your trade exits.
Up Trending Market with Low Volatility
Up Trending Market with High Volatility
Sideways Market with Low Volatility
Sideways Market with High Volatility
Down Trending Market with Low Volatility
Down Trending Market with High Volatility
Now determine which indicator works best with a certain stock and in a particular market condition. It’s almost impossible to test all of the different possibilities. Add to that the amount of time reviewing charts and making adjustments to indicators for changing market conditions. Too many hours required, I have better things to do every evening that scanning, reviewing and updating charts all night.
This is where Neural Network Technology comes in to play. This software takes all the trade information over a given period (O,H,L,C,V), adds to that many technical indicators and calculates deltas for data input. After downloading the historical data, it then learns and looks for patterns in the stock depending on thousands of combinations of the input data. It ignores which inputs do not have a correlation to the output and then uses the inputs which directly affect the outcome of the model. The software then generates signals for buys, sells, holds or neutral from this learned process. Will this system give you 100% accurate trades, absolutely not! Will it give you an edge over the traditional way of using technical analysis and save you time, YES!
As you know, institutional money managers, hedge funds and large buyers/sellers are truly what move a stock price. You can find great value companies with strong fundamentals, but the stock goes nowhere unless big money is interested. In past years and increasingly more every day, these players have been utilizing software (Algorithmic Trading) to gain an advantage over other traders. Here is an example. A stock has bottomed out and a large player has taken a position long. Usually most retail and smaller traders will not take a position until a technical indicator has given them a signal such as a 20/50 simple moving average crossover or other lagging indicator. The stock begins trending up but is a ways from the crossover and begins to stall. The large player does not see a catalyst for moving the stock higher as they originally thought. Depending on volume, share float and market conditions, an algo could purchase enough stock to push the price higher, triggering the crossover. This would then signal additional buying of the stock from other smaller traders. The algo then knows with these new smaller buyers entering the market that it can begin to sell the stock without creating too much downside pressure. Now two situations usually occur, neither good for the smaller buyer of the stock. The algo will continue to sell the stock as long as small buyers are still there near that price. If the large player can unload all of his shares here, they are done and out of the trade. With no large buyers in the stock anymore, the stock just sits there and the small traders are left with a stock that goes nowhere. The second thing that happens is the worst scenario for the smaller trader. If there are not enough small buyers at this price, the seller of the stock now must sell the shares at a lower cost to unwind their position. EXIT THE TRADE, the reason for being in the trade is now gone. The algo will do this at a pace not to cause panic selling until the position is closed for the large player. The smaller buyers of the stock now are holding a losing position and some may double up and buy additional shares (WRONG). The large player goes away with a small winner and the smaller trader is handed another loser. Now what if, the stock had showed this pattern before not holding support after crossover. Would you have spotted this on the chart with a certain indicator, probably not? Could Neural Network Technology learned this move and steered you clear of this trade, possibly yes. I would rather have this extra edge than go without.
To be a successful trader or investor, your losers must be smaller than your winners and it’s the trades you decide to skip because of no edge that allow you to stay in the game. The example above is just one of the many battles each day in the war of trading and investing.